library(Seurat)
library(ggplot2)
library(RColorBrewer)
library(writexl)
library(tidyverse)
color <- c(brewer.pal(8,"Paired"), brewer.pal(8,"Set3"),
brewer.pal(8,"Set2"), brewer.pal(8,"Accent"),
brewer.pal(9,"Pastel1"))
data2 <- readRDS("/Users/xu/Desktop/CART/data/qcdata/batch2_afterqcdata.rds")
spe <- readRDS("/Users/xu/Desktop/CART/data/qcdata/2025-02-21_sfe_qc.rds")
seurat <- readRDS("/Users/xu/Desktop/CART/data/seuratrawdata/batch1_CARTXenium_seurat_250218130928.rds")
dim(spe@colData)
## [1] 50042 21
cellid <- rownames(spe@colData)
data1<- seurat[,cellid]
merged_obj <- merge(data1, y = data2)
merged_obj$batch <- rep(c("batch1", "batch2"), times = c(ncol(data1), ncol(data2)))
merged_obj <- JoinLayers(merged_obj)
merged_obj <- NormalizeData(merged_obj,scale.factor = 300)
all.genes <- rownames(merged_obj)
merged_obj<- ScaleData(merged_obj ,features = all.genes, split.by=merged_obj@meta.data$SampleId)
merged_obj <- RunPCA(merged_obj,features = all.genes)
ElbowPlot(merged_obj,ndims=50,reduction = "pca")
merged_obj <- FindNeighbors(merged_obj,dims =1:30)
merged_obj <- FindClusters(merged_obj,resolution = 1, algorithm = 4)
merged_obj <- RunUMAP(merged_obj, dims = 1:30)
DimPlot(merged_obj, reduction ="umap",pt.size = 0.25,alpha = 0.2, label = TRUE, label.size = 2, repel = TRUE,cols = color)
for (cluster_id in unique(merged_obj@meta.data$seurat_clusters)) {
cluster_subset <- subset(merged_obj, seurat_clusters == cluster_id)
assign(paste0("c", cluster_id), cluster_subset)
}
c6<- FindClusters(c6, resolution = 0.8)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 3834
## Number of edges: 119401
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.7427
## Number of communities: 9
## Elapsed time: 0 seconds
c6$annotation <- "Unknown"
c6$annotation[c6$seurat_clusters%in%c(4)] <- "fibroblast"
c6$annotation[c6$seurat_clusters%in%c(3,5)] <- "macrophage"
c6$annotation[c6$seurat_clusters%in%c(0,1,2,6,7,8)] <-"T cell"
c11 <- FindClusters(c11, resolution = 0.8)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 2821
## Number of edges: 95849
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.7379
## Number of communities: 5
## Elapsed time: 0 seconds
c11$annotation <- "Unknown"
c11$annotation[c11$seurat_clusters%in%c(1,3)] <- "fibroblast"
c11$annotation[c11$seurat_clusters%in%c(0,2,4)] <-"smooth muscle cell"
c13 <- FindClusters(c13, resolution = 0.8)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 2418
## Number of edges: 99415
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.6947
## Number of communities: 6
## Elapsed time: 0 seconds
c13$annotation <- "Unknown"
c13$annotation[c13$seurat_clusters==5] <- "smooth muscle cell"
c13$annotation[c13$seurat_clusters%in%c(0,1,2,3,4)] <-"keratinocyte"
c18 <- FindClusters(c18, resolution = 0.8)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 2233
## Number of edges: 58298
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.7666
## Number of communities: 10
## Elapsed time: 0 seconds
c18$annotation <- "Unknown"
c18$annotation[c18$seurat_clusters==8] <- "fibroblast"
c18$annotation[c18$seurat_clusters==3] <- "T cell"
c18$annotation[c18$seurat_clusters%in%c(0,1,2,4,5,6,7,9)] <-"keratinocyte"
c20 <- FindClusters(c20, resolution = 0.8)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 1090
## Number of edges: 31713
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.6142
## Number of communities: 5
## Elapsed time: 0 seconds
c20$annotation <- "Unknown"
c20$annotation[c20$seurat_clusters==0] <- "fibroblast"
c20$annotation[c20$seurat_clusters%in%c(1,2,3,4)] <- "T cell"
c22 <- FindClusters(c22, resolution = 0.8)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 993
## Number of edges: 35865
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.6805
## Number of communities: 5
## Elapsed time: 0 seconds
c22$annotation <- "Unknown"
c22$annotation[c22$seurat_clusters%in%c(1,3)] <- "melanocyte"
c22$annotation[c22$seurat_clusters%in%c(0,2,4)] <- "keratinocyte"
c25 <- FindClusters(c25, resolution = 0.8)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 748
## Number of edges: 27468
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.5710
## Number of communities: 4
## Elapsed time: 0 seconds
c25$annotation <- "Unknown"
c25$annotation[c25$seurat_clusters==2] <- "fibroblast"
c25$annotation[c25$seurat_clusters%in%c(0,1,3)] <- "Schwann cell"
c31 <- FindClusters(c31, resolution = 0.8)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 267
## Number of edges: 6423
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.5758
## Number of communities: 4
## Elapsed time: 0 seconds
c31$annotation <- "Unknown"
c31$annotation[c31$seurat_clusters%in%c(0,2)] <- "keratinocyte"
c31$annotation[c31$seurat_clusters%in%c(1,3)] <- "Langerhans cell"
merged_obj$first_annotation <- "unknown"
merged_obj$first_annotation[merged_obj$seurat_clusters%in%c(4,7,8,12,14,15,17,19,21,24,27,30,32)] <- "keratinocyte"
merged_obj$first_annotation[merged_obj$seurat_clusters%in%c(2,3,9,16)] <- "fibroblast"
merged_obj$first_annotation[merged_obj$seurat_clusters%in%c(5,28)] <- "endothelial cell"
merged_obj$first_annotation[merged_obj$seurat_clusters%in%c(29)] <- "adipocyte"
merged_obj$first_annotation[merged_obj$seurat_clusters%in%c(1,26)] <- "T cell"
merged_obj$first_annotation[merged_obj$seurat_clusters%in%c(23)] <- "mast cell"
merged_obj$first_annotation[merged_obj$seurat_clusters%in%c(10)] <- "macrophage"
merged_obj$first_annotation<- c31$annotation
merged_obj$first_annotation <- c25$annotation
merged_obj$first_annotation <- c22$annotation
merged_obj$first_annotation<- c20$annotation
merged_obj$first_annotation<- c18$annotation
merged_obj$first_annotation <- c13$annotation
merged_obj$first_annotation <- c11$annotation
merged_obj$first_annotation <- c6$annotation
table(merged_obj$first_annotation)
##
## adipocyte endothelial cell fibroblast keratinocyte
## 355 4285 19123 30562
## Langerhans cell macrophage mast cell melanocyte
## 119 4033 948 427
## Schwann cell smooth muscle cell T cell
## 571 1944 12269
DimPlot(merged_obj, reduction ="umap",pt.size = 0.25,alpha = 0.2 ,label = TRUE, label.size = 2, repel = TRUE,cols = color,group.by = "first_annotation")
merged_obj$major_celltype <- merged_obj$first_annotation
merged_obj$major_celltype[merged_obj$major_celltype%in%c("Langerhans cell","macrophage","mast cell","T cell")] <- "immune cell"
table(merged_obj$major_celltype)
##
## adipocyte endothelial cell fibroblast immune cell
## 355 4285 19123 17369
## keratinocyte melanocyte Schwann cell smooth muscle cell
## 30562 427 571 1944
DimPlot(merged_obj, reduction ="umap",pt.size = 0.25,alpha = 0.2, label = TRUE, label.size = 2, repel = TRUE,cols = color,group.by = "major_celltype")
merged_obj$Patient[merged_obj$SampleId=="SSc213"] <- "Patient2"
merged_obj$fourgroups <- merged_obj$Group
merged_obj$fourgroups [merged_obj$Group%in%c("SSc_postCART_1m","SSc_postCART_2m")] <- "SSc_postCART_early"
merged_obj$fourgroups [merged_obj$Group%in%c("SSc_postCART_4m","SSc_postCART_6m")] <- "SSc_postCART_late"
merged_obj$fourgroups [merged_obj$Group%in%c("SSc_postCART_9m","SSc_postCART_1y")] <- "SSc_postCART_superlate"
table(merged_obj$fourgroups)
##
## SSc_postCART_early SSc_postCART_late SSc_postCART_superlate
## 25178 13618 12053
## SSc_preCART
## 23787
saveRDS(merged_obj,"/Users/xu/Desktop/1_project/merged_obj_manually.rds")
merged_obj <- readRDS("/Users/xu/Desktop/1_project/merged_obj_manually.rds")
autodata <- readRDS("/Users/xu/Desktop/CART/data/annotation/major celltype/mergeddata/Erlangen_CART_XeniumScimilarity250320112810.rds")
DimPlot(autodata, reduction ="umap.unintegrated", pt.size = 0.25, alpha=0.2,label = TRUE, label.size = 5, repel = TRUE,,group.by="scimilarity_hint")
table(autodata$scimilarity_hint)
##
## B cell CD4-positive, alpha-beta T cell
## 379 4371
## CD8-positive, alpha-beta T cell Langerhans cell
## 1179 45
## basal cell blood vessel endothelial cell
## 23310 27
## capillary endothelial cell classical monocyte
## 3427 1819
## endothelial cell of lymphatic vessel erythrocyte
## 444 4
## fibroblast glial cell
## 19761 307
## innate lymphoid cell keratinocyte
## 45 3548
## macrophage mast cell
## 5367 835
## melanocyte myofibroblast cell
## 628 3224
## natural killer cell neutrophil
## 666 359
## plasma cell plasmacytoid dendritic cell
## 799 1011
## platelet progenitor cell
## 25 869
## regulatory T cell smooth muscle cell
## 1178 1009
unique(autodata@meta.data$scimilarity_hint)
## [1] basal cell keratinocyte
## [3] melanocyte macrophage
## [5] fibroblast endothelial cell of lymphatic vessel
## [7] myofibroblast cell mast cell
## [9] capillary endothelial cell classical monocyte
## [11] regulatory T cell progenitor cell
## [13] CD4-positive, alpha-beta T cell Langerhans cell
## [15] platelet smooth muscle cell
## [17] glial cell neutrophil
## [19] natural killer cell B cell
## [21] plasma cell blood vessel endothelial cell
## [23] plasmacytoid dendritic cell CD8-positive, alpha-beta T cell
## [25] innate lymphoid cell erythrocyte
## 26 Levels: B cell ... smooth muscle cell
category_mapping <- c(
"basal cell" = "keratinocyte",
"keratinocyte" = "keratinocyte",
"fibroblast" = "fibroblast",
"macrophage" = "immune cell",
"classical monocyte" = "immune cell",
"regulatory T cell" = "immune cell",
"CD4-positive, alpha-beta T cell" = "immune cell",
"CD8-positive, alpha-beta T cell" = "immune cell",
"Langerhans cell" = "immune cell",
"mast cell" = "immune cell",
"natural killer cell" = "immune cell",
"B cell" = "immune cell",
"plasma cell" = "immune cell",
"neutrophil" = "immune cell",
"innate lymphoid cell" = "immune cell",
"plasmacytoid dendritic cell" = "immune cell",
"platelet" = "platelet",
"progenitor cell" = "progenitor cell",
"erythrocyte" = "erythrocyte",
"myofibroblast cell" = "fibroblast",
"capillary endothelial cell" = "endothelial cell",
"blood vessel endothelial cell" = "endothelial cell",
"endothelial cell of lymphatic vessel" = "endothelial cell",
"glial cell" = "Schwann cell",
"melanocyte" = "melanocyte",
"smooth muscle cell" = "smooth muscle cell",
"vascular associated smooth muscle cell" = "smooth muscle cell"
)
autodata$scimilarity_hint <- as.character(autodata$scimilarity_hint)
category_mapping <- unlist(category_mapping)
autodata@meta.data$major_scimilarity_hint <- category_mapping[autodata@meta.data$scimilarity_hint]
autodata@meta.data$major_scimilarity_hint[is.na(autodata@meta.data$major_scimilarity_hint)] <- "unknown"
table(autodata@meta.data$major_scimilarity_hint)
##
## endothelial cell erythrocyte fibroblast immune cell
## 3898 4 22985 18053
## keratinocyte melanocyte platelet progenitor cell
## 26858 628 25 869
## Schwann cell smooth muscle cell
## 307 1009
DimPlot(autodata, reduction ="umap.unintegrated", pt.size = 0.25, label = TRUE, label.size = 5, repel = TRUE,,group.by="major_scimilarity_hint")
autodata$major_celltype <- merged_obj$major_celltype
table(autodata$major_celltype)
##
## adipocyte endothelial cell fibroblast immune cell
## 355 4285 19123 17369
## keratinocyte melanocyte Schwann cell smooth muscle cell
## 30562 427 571 1944
diff_annotations <- autodata$major_celltype!= autodata$major_scimilarity_hint
data_diff <- subset(autodata, cells = which(diff_annotations))
table(diff_annotations)
## diff_annotations
## FALSE TRUE
## 63725 10911
#FALSE TRUE
#63725 10911
autodata$highlight <- ifelse(diff_annotations, "differ", "same")
DimPlot(autodata, group.by = "highlight") +
ggtitle("UMAP with differing annotations") +
scale_color_manual(values = c("#866aa3", "grey")) +
theme_minimal()+
theme(
panel.grid = element_blank(),
plot.title = element_text(face = "bold", size = 16)
)
all.genes <- rownames(data_diff)
data_diff <- RunPCA(data_diff,features = all.genes)
ElbowPlot(data_diff,ndims=50,reduction = "pca")
data_diff <- FindNeighbors(data_diff,dims =1:30)
data_diff <- FindClusters(data_diff,resolution = 1, algorithm = 4)
data_diff <- RunUMAP(data_diff, dims = 1:30)
DimPlot(data_diff, reduction ="umap", pt.size = 0.25, label = TRUE, label.size = 3, repel = TRUE,cols = color)
for (cluster_id in unique(data_diff@meta.data$seurat_clusters)) {
cluster_subset <- subset(data_diff, seurat_clusters == cluster_id)
assign(paste0("cl", cluster_id), cluster_subset)
}
cl1<- FindClusters(cl1, resolution = 0.8)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 1192
## Number of edges: 34432
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.6865
## Number of communities: 7
## Elapsed time: 0 seconds
cl1$finalannotation <- "Unknown"
cl1$finalannotation[cl1$seurat_clusters%in%c(0,6)] <- "fibroblast"
cl1$finalannotation[cl1$seurat_clusters%in%c(1,2,3,4,5)] <- "smooth muscle cell"
table(cl1$finalannotation)
##
## fibroblast smooth muscle cell
## 448 744
cl2<- FindClusters(cl2, resolution = 0.8)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 1163
## Number of edges: 54447
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.5755
## Number of communities: 5
## Elapsed time: 0 seconds
cl2$finalannotation <- "Unknown"
cl2$finalannotation[cl2$seurat_clusters%in%c(1,2,4)] <- "fibroblast"
cl2$finalannotation[cl2$seurat_clusters%in%c(0,3)] <- "keratinocyte"
table(cl2$finalannotation)
##
## fibroblast keratinocyte
## 674 489
cl3<- FindClusters(cl3, resolution =1.2)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 1063
## Number of edges: 39495
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.5850
## Number of communities: 8
## Elapsed time: 0 seconds
cl3$finalannotation <- "Unknown"
cl3$finalannotation[cl3$seurat_clusters%in%c(0,2,3)] <- "fibroblast"
cl3$finalannotation[cl3$seurat_clusters%in%c(1,4,5,6,7)] <- "immune cell"
table(cl3$finalannotation)
##
## fibroblast immune cell
## 493 570
cl5<- FindClusters(cl5, resolution =1.2)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 891
## Number of edges: 34710
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.4978
## Number of communities: 7
## Elapsed time: 0 seconds
cl5$finalannotation <- "Unknown"
cl5$finalannotation[cl5$seurat_clusters%in%c(3,5,6)] <- "keratinocyte"
cl5$finalannotation[cl5$seurat_clusters%in%c(0,1,2,4)] <- "adipocyte"
table(cl5$finalannotation)
##
## adipocyte keratinocyte
## 685 206
cl6<- FindClusters(cl6, resolution =1.2)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 883
## Number of edges: 33655
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.4666
## Number of communities: 6
## Elapsed time: 0 seconds
cl6$finalannotation <- "Unknown"
cl6$finalannotation[cl6$seurat_clusters%in%c(1,2,3,4,5)] <- "immune cell"
cl6$finalannotation[cl6$seurat_clusters%in%c(0)] <- "fibroblast"
table(cl6$finalannotation)
##
## fibroblast immune cell
## 203 680
cl9<- FindClusters(cl9, resolution =1.2)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 554
## Number of edges: 16441
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.4358
## Number of communities: 7
## Elapsed time: 0 seconds
cl9$finalannotation <- "Unknown"
cl9$finalannotation[cl9$seurat_clusters%in%c(1,3)] <- "fibroblast"
cl9$finalannotation[cl9$seurat_clusters%in%c(0,2,4,5,6)] <- "endothelial cell"
table(cl9$finalannotation)
##
## endothelial cell fibroblast
## 343 211
cl10<- FindClusters(cl10, resolution =1.2)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 518
## Number of edges: 17434
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.4674
## Number of communities: 6
## Elapsed time: 0 seconds
cl10$finalannotation <- "Unknown"
cl10$finalannotation[cl10$seurat_clusters%in%c(1,2,3,4,5)] <- "fibroblast"
cl10$finalannotation[cl10$seurat_clusters%in%c(0)] <- "smooth muscle cell"
table(cl10$finalannotation)
##
## fibroblast smooth muscle cell
## 414 104
cl13<- FindClusters(cl13, resolution =1.2)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 353
## Number of edges: 8945
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.5956
## Number of communities: 5
## Elapsed time: 0 seconds
cl13$finalannotation <- "Unknown"
cl13$finalannotation[cl13$seurat_clusters%in%c(0,4)] <- "smooth muscle cell"
cl13$finalannotation[cl13$seurat_clusters%in%c(1,2,3)] <- "keratinocyte"
table(cl13$finalannotation)
##
## keratinocyte smooth muscle cell
## 183 170
cl16<- FindClusters(cl16, resolution =1.2)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 185
## Number of edges: 6468
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.2701
## Number of communities: 6
## Elapsed time: 0 seconds
cl16$finalannotation <- "Unknown"
cl16$finalannotation[cl16$seurat_clusters%in%c(0,1,2,4,5)] <- "immune cell"
cl16$finalannotation[cl16$seurat_clusters%in%c(3)] <- "keratinocyte"
table(cl16$finalannotation)
##
## immune cell keratinocyte
## 153 32
cl17<- FindClusters(cl17, resolution =1.2)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 179
## Number of edges: 5937
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.3225
## Number of communities: 4
## Elapsed time: 0 seconds
cl17$finalannotation <- "Unknown"
cl17$finalannotation[cl17$seurat_clusters%in%c(1)] <- "fibroblast"
cl17$finalannotation[cl17$seurat_clusters%in%c(0,2,3)] <- "smooth muscle cell"
table(cl17$finalannotation)
##
## fibroblast smooth muscle cell
## 59 120
cl18<- FindClusters(cl18, resolution =1.2)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 132
## Number of edges: 2800
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.4179
## Number of communities: 4
## Elapsed time: 0 seconds
cl18$finalannotation <- "Unknown"
cl18$finalannotation[cl18$seurat_clusters%in%c(0,1,3)] <- "fibroblast"
cl18$finalannotation[cl18$seurat_clusters%in%c(2)] <- "keratinocyte"
table(cl18$finalannotation)
##
## fibroblast keratinocyte
## 105 27
cl19<- FindClusters(cl19, resolution =1.2)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 112
## Number of edges: 2787
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.1852
## Number of communities: 6
## Elapsed time: 0 seconds
cl19$finalannotation <- "Unknown"
cl19$finalannotation[cl19$seurat_clusters%in%c(0,3)] <- "fibroblast"
cl19$finalannotation[cl19$seurat_clusters%in%c(1,2,4)] <- "immune cell"
table(cl19$finalannotation)
##
## fibroblast immune cell
## 53 59
data_diff$finalannotation <- "unkown"
data_diff$finalannotation[data_diff$seurat_clusters %in%c(4,7,8,14,15)] <-"keratinocyte"
data_diff$finalannotation[data_diff$seurat_clusters== 12] <-"adipocyte"
data_diff$finalannotation[data_diff$seurat_clusters==11] <-"Schwann cell"
data_diff$finalannotation <-cl1$finalannotation
data_diff$finalannotation <-cl2$finalannotation
data_diff$finalannotation <-cl3$finalannotation
data_diff$finalannotation<-cl5$finalannotation
data_diff$finalannotation<-cl6$finalannotation
data_diff$finalannotation<-cl9$finalannotation
data_diff$finalannotation<-cl10$finalannotation
data_diff$finalannotation <-cl13$finalannotation
data_diff$finalannotation <-cl16$finalannotation
data_diff$finalannotation<-cl17$finalannotation
data_diff$finalannotation<-cl18$finalannotation
data_diff$finalannotation <-cl19$finalannotation
table(data_diff$finalannotation)
##
## adipocyte endothelial cell fibroblast immune cell
## 1041 343 2660 1462
## keratinocyte Schwann cell smooth muscle cell
## 3895 372 1138
DimPlot(data_diff, reduction ="umap", pt.size = 0.25, label = TRUE, label.size = 5, repel = TRUE,cols = color,group.by="finalannotation")
####
reannotated_data <- data_diff@meta.data$finalannotation
names(reannotated_data) <- rownames(data_diff@meta.data)
matching_cells <- intersect(rownames(autodata@meta.data), names(reannotated_data))
identical(names(reannotated_data[matching_cells]), matching_cells)
## [1] TRUE
autodata$major_celltype[rownames(autodata@meta.data) %in% matching_cells] <- reannotated_data[matching_cells]
table(autodata$major_celltype)
##
## adipocyte endothelial cell fibroblast immune cell
## 1041 3828 19878 16917
## keratinocyte melanocyte Schwann cell smooth muscle cell
## 30359 370 633 1610
merged_obj $merged_annotation<- autodata$major_celltype
table(merged_obj$merged_annotation)
##
## adipocyte endothelial cell fibroblast immune cell
## 1041 3828 19878 16917
## keratinocyte melanocyte Schwann cell smooth muscle cell
## 30359 370 633 1610
table(merged_obj$first_annotation)
##
## adipocyte endothelial cell fibroblast keratinocyte
## 355 4285 19123 30562
## Langerhans cell macrophage mast cell melanocyte
## 119 4033 948 427
## Schwann cell smooth muscle cell T cell
## 571 1944 12269
DimPlot(merged_obj, reduction ="umap", pt.size = 0.25, label = TRUE, label.size = 5, repel = TRUE,group.by="merged_annotation")
#
cladipo <- subset(merged_obj,merged_annotation=="adipocyte")
cladipo<- FindClusters(cladipo, resolution =1.2)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 1041
## Number of edges: 32269
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.6624
## Number of communities: 12
## Elapsed time: 0 seconds
cladipo$finalannotation <- "Unknown"
cladipo$finalannotation[cladipo$seurat_clusters%in%c(0,1,3)] <- "keratinocyte"
cladipo$finalannotation[cladipo$seurat_clusters%in%c(2,4,5)] <- "adipocyte"
table(cladipo$finalannotation)
##
## adipocyte keratinocyte
## 361 680
merged_obj$merged_annotation <- cladipo$finalannotation
table(merged_obj$merged_annotation)
##
## adipocyte endothelial cell fibroblast immune cell
## 361 3828 19878 16917
## keratinocyte melanocyte Schwann cell smooth muscle cell
## 31039 370 633 1610
####
clsmc <- subset(merged_obj,merged_annotation=="smooth muscle cell")
clsmc<- FindClusters(clsmc, resolution =1.2)
## Modularity Optimizer version 1.3.0 by Ludo Waltman and Nees Jan van Eck
##
## Number of nodes: 1610
## Number of edges: 45980
##
## Running Louvain algorithm...
## Maximum modularity in 10 random starts: 0.6726
## Number of communities: 26
## Elapsed time: 0 seconds
clsmc$finalannotation <- "Unknown"
clsmc$finalannotation[clsmc$seurat_clusters%in%c(7)] <- "fibroblast"
clsmc$finalannotation[clsmc$seurat_clusters%in%c(6)] <- "pericyte"
clsmc$finalannotation[clsmc$seurat_clusters%in%c(8,10)] <- "immune cell"
clsmc$finalannotation[clsmc$seurat_clusters%in%c(0,1,2,3,4,5,9)] <- "smooth muscle cell"
table(clsmc$finalannotation)
##
## fibroblast immune cell pericyte smooth muscle cell
## 97 5 101 1407
merged_obj$merged_annotation<- clsmc$finalannotation
table(merged_obj$merged_annotation)
##
## adipocyte endothelial cell fibroblast immune cell
## 361 3828 19975 16922
## keratinocyte melanocyte pericyte Schwann cell
## 31039 370 101 633
## smooth muscle cell
## 1407
#############
merged_obj$merged_annotation[merged_obj$merged_annotation=="melanocyte"] <- "keratinocyte"
table(merged_obj$merged_annotation)
##
## adipocyte endothelial cell fibroblast immune cell
## 361 3828 19975 16922
## keratinocyte pericyte Schwann cell smooth muscle cell
## 31409 101 633 1407
DimPlot(merged_obj, reduction ="umap", pt.size = 0.25, label = TRUE, label.size = 5, repel = TRUE,cols = color,group.by="merged_annotation")
majormarkergene <- c("KRT5","KRT19","KRT20",
"COL1A1","COL1A2","LUM",
"VWF","ERG","PECAM1",
"CD3E","CD68","CD19",
"MLANA","KIT","MITF",
"MYL9","TAGLN","MYH11",
"SOX10","GLDN","MPZ",
"PDGFRA","PDGFRB","RGS5",
"PPARG","ADIPOQ","FABP4 ","PLIN1", "PLIN4"
)
metadata <- merged_obj@meta.data
color <-c('#E5D2DD', '#53A85F')
p <- DotPlot(merged_obj, features = unique(majormarkergene),cols = color, group.by = "merged_annotation"#, split.by = "group"
)+coord_flip()
exp <- p$data
exp$features.plot <- as.factor(exp$features.plot)
exp$features.plot <- fct_inorder(exp$features.plot)
exp$id <- as.factor(exp$id)
exp$id <- fct_inorder(exp$id)
colnames(exp)[5] <- 'AverageExpression'
colnames(exp)[2] <- 'Percent'
cell_types <- levels(exp$id)
xintercepts <- seq(1.5, length(cell_types) - 0.5, by=1)
ggplot(exp, aes(x=id, y=features.plot))+
geom_point(aes(size= Percent, color=AverageExpression))+
geom_point(aes(size=Percent, color=AverageExpression), shape=21, color="black", stroke=1)+
scale_size_continuous(range = c(2,6))+
theme(panel.grid = element_blank(),
axis.line = element_blank(),
axis.ticks = element_blank(),
panel.background = element_rect(color = 'black', size = 1.2, fill = 'transparent'),
axis.text.x = element_text(size=11, color="black", angle=45, hjust = 1),
axis.text.y = element_text(size=11, color="black"))+
scale_color_gradientn(colors = colorRampPalette(c("#c1719c","#d9bad0","#d0ecd7","#a3e0bb","#4fc3b0"))(100))+
labs(x=NULL, y=NULL)+
geom_vline(xintercept = xintercepts, linetype="dotted", size=0.8)
merged_obj$automaticalannotation <- autodata$scimilarity_hint
saveRDS(merged_obj,"/Users/xu/Desktop/1_project/merged_obj.rds")